【作者】 刘鹏；

【导师】 朱齐丹；

【作者基本信息】 哈尔滨工程大学 ， 控制科学与工程， 2017， 硕士

【摘要】 智能移动机器人在工业生产、家庭服务和医疗保健等领域都有着广阔的应用前景,机器人的自主导航能力是其智能化的基础。本文针对机器人的室内定位、全局、局部路径规划和路径跟踪技术进行研究,具体的研究内容包括:1.介绍了常用的人工路标识别算法—基于颜色特征识别。针对这种识别算法不足,提出了一种对光照和环境变化具有良好鲁棒性的识别方法:改进的基于SURF特征的路标识别算法。使用全景视觉系统获得环境信息,首先将特征匹配算法应用到人工路标的识别中,针对该算法设计了利于识别和定位的路标;然后利用设置动态检测窗口的方法缩小了路标检测区域,解决了全景图像的特征提取运算量大的问题,提高了算法识别率;接下来采用路标模版实时更新的策略,消除了视角变化对识别率的影响。最后介绍了基于人工路标的定位方法。多个场景下的对比实验证明了本文的路标识别算法相比于颜色识别算法具有更强的鲁棒性。2.介绍了地图模型的建立方法,建立了栅格地图。介绍了经典的全局路径规划方法—A*算法。分析了算法在应用于移动机器人路径规划中缺乏安全性的不足,在搜索中增加了约束条件,提高了机器人的安全性能。针对A*算法路径不够平滑的缺点,在不增加栅格密度的条件下提出了路径优化算法,可以有效缩短路径的距离、减少转弯角度,降低了机器人的能耗。最后的多个情况下的对比实验验证了本文提出算法的有效性。3.建立了机器人的运动模型,分析了本文需要实现的跟踪和局部避障的具体功能,确定了先全局规划再局部导航的分层控制策略:首先进行全局路径规划,设置局部目标,将全局路径分段;然后依次对局部目标进行导航、同时避障,简化了问题。使用模糊控制算法进行机器人局部路径跟踪和避障,介绍了模糊控制器的基本概念和设计过程,最后针对本文使用的移动机器人设计了模糊控制器。4.介绍了综合实验的环境和硬件平台,编写了集运动控制、远程通讯、路标识别、室内定位、全局路径规划、轨迹跟踪与避碰等功能于一身的软件成果。设计了机器人室内定位实验,确定了三路标下的定位范围与定位精度。最后设计了综合实验,验证了定位、全局路径规划、局部避碰和路径跟踪算法的有效性。更多还原

【Abstract】 Intelligent mobile robot has broad application prospects in industrial production,family services,medical services and other fields,and the autonomous navigation ability of robot is the foundation of its intelligence.Aiming at this issue,indoor positioning,path planning and path tracking algorithms have been studied.Specific research contents are as follows:1.The commonly used algorithm of artificial landmark recognition-color feature based algorithm has been introduced.Aimed at the deficiency of this algorithm,a rapid localization method was proposed,which has good robustness to illumination and environmental changes.Using omni-directional vision system to obtain environmental information,Firstly,the feature matching algorithm is applied to the landmark recognition.Then the landmark is designed which is conducive to recognition and localization according to this algorithm.Secondly,setting dynamic detection window to reduce the landmark detection area and solve the problem caused by the large amount of computation,improved the recognition rate.After that the affection of the angle of view on the recognition rate is eliminated by updating templet in real time.Finally,the localization algorithm based on the artificial landmark is introduced.Many experiment under different scenarios indicate that this method is more robust than color recognition.2.The establishment method of map model is introduced,and the grid map of experimental environment is established.The classical global path planning method : A*algorithm is introduced in this paper.Aimed at the deficiency of this algorithm,an improved algorithm is proposed to improve security of robot.And the path optimization algorithm is proposed as well,under the condition of not increasing the density of the grid.In most cases,the distance of the path can be reduced effectively,the turning angle can be reduced,and the energy consumption of the robot can be reduced.Finally,the effectiveness of the algorithm is verified by the contrast experiment.3.Secondly,the motion model of the mobile robot is established.According to need of the path tracking and obstacle avoidance algorithms,the global-local control strategy is proposed.The fuzzy controller is used to solve this problem.The basic concept and design process of the fuzzy controller are introduced.Finally,the fuzzy controller is designed for the mobile robot used in this paper.4.The environment and hardware platform of comprehensive experiment are introduced,as well as the software which has motion control,remote communication,landmarkrecognition,indoor positioning,global path planning,trajectory tracking and collision avoidance functions.The robot indoor positioning experiment was designed to determine the positioning range and positioning accuracy under three landmarks.Finally,a comprehensive experiment was designed to verify the effectiveness of all proposed algorithms.更多还原